mlsanityData to 14 Oct 2026

Harmonizing Spectral Evolution in Conditional Flow Matching for TTS

arXiv 2609.34431 ↗First posted 28 Sep 2026cs.LG cs.SDcited by 0

On the map: Diffusion models → Audio, speech and music

Abstract (arXiv)

Conditional Flow Matching (CFM) models for text-to-speech (TTS) suffer from incoherent frequency evolution during inference. While similar spectral imbalances are addressed in diffusion models for other domains, those generic solutions fail to generalize to the inherently uncoordinated acoustic dynamics of CFM. We demonstrate that this issue can be effectively mitigated by introducing a novel training-free frequency-selective boosting strategy. Using the Discrete Wavelet Transform (DWT), our method dynamically modulates mel-spectrogram sub-bands during ODE integration, synchronizing spectral development by penalizing aggressive low-frequency growth and boosting lagging high-frequency details. Validated across diverse architectures (Matcha-TTS, F5-TTS, IndicF5), our approach reduces the required Number of Function Evaluations (NFE) from 32 to 26 and improves Frechet Audio Distance (FAD) by up to 61%, all without compromising mean opinion scores, speaker similarity, and speech intelligibility.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Audio, speech and music, one of 14 directions in Diffusion models. Builds on 15 of its 19 references that are in the map. First posted on arXiv 28 Sep 2026.

Builds on

15 of 19 references.

Closest ancestors · 2

Same direction, newest first

  1. 2024F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching9 Oct
  2. 2023Matcha-TTS: A fast TTS architecture with conditional flow matching6 Sep

Other references · 13

  1. 2022Flow Matching for Generative Modelingcited by 5,702
  2. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 4,057
  3. 2018Neural Ordinary Differential Equationscited by 3,779
  4. 2018On the Spectral Bias of Neural Networkscited by 1,341
  5. 2016CNN Architectures for Large-Scale Audio Classificationcited by 704
  6. 2019LibriTTS: A Corpus Derived from LibriSpeech for Text-to-Speechcited by 649
  7. 2022UTMOS: UTokyo-SaruLab System for VoiceMOS Challenge 2022cited by 390
  8. 2023Voicebox: Text-Guided Multilingual Universal Speech Generation at Scalecited by 379
  9. 2018Fr'echet Audio Distance: A Metric for Evaluating Music Enhancement Algorithmscited by 225
  10. 2022Diffusion Probabilistic Model Made Slimcited by 121
  11. 2024Boosting Diffusion Models with Moving Average Sampling in Frequency Domaincited by 55
  12. 2023Vistaar: Diverse Benchmarks and Training Sets for Indian Language ASRcited by 21
1 more
  1. 2024IndicVoices-R: Unlocking a Massive Multilingual Multi-speaker Speech Corpus for Scaling Indian TTScited by 16

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Towards Flow-Matching-based TTS without Classifier-Free Guidance
  2. 2026Beyond Two-stage Diffusion TTS: Joint Structure and Content Refinement via Jump Diffusion
  3. 2025DiFlow-TTS: Compact and Low-Latency Zero-Shot Text-to-Speech with Discrete Flow Matching
  4. 2025Shallow Flow Matching for Coarse-to-Fine Text-to-Speech Synthesis
  5. 2025Accelerating Flow-Matching-Based Text-to-Speech via Empirically Pruned Step Sampling
  6. 2025FlowSE: Efficient and High-Quality Speech Enhancement via Flow Matching
  7. 2024DPI-TTS: Directional Patch Interaction for Fast-Converging and Style Temporal Modeling in Text-to-Speech
  8. 2026ARCHI-TTS: A flow-matching-based Text-to-Speech Model with Self-supervised Semantic Aligner and Accelerated Inference